AI Implementation Risks and Governance in Healthcare

Hey L&D Managers -- it's up to you to make your case to your bosses. All the way up to the CFO if necessary. Read Josh Cavalier's post below for an eye-opening reality check that some of you may already be living. Let's start a conversation around how to talk to your company's leadership about how to do AI right, and the consequences of doing it wrong.

I love what Josh has to say, here. AI can be a double-edged sword, but the key is to use it wisely and be aware of those edge cases and scenarios where you - or a role/function, etc. might be at risk so you can mitigate the potential impacts.

"AI can't do your job, but an AI salesman can convince your boss to fire you and replace you with an AI that can't do your job." Cory Doctorow

This weekend, I came across a Reddit thread titled "So sick of AI" that gave me pause. I had to read it over a few times. An instructional designer described their company's "AI first" mandate: consequences for not using AI, leadership openly stating their goal is to replace designers, one ID already let go because "AI can handle that 10%." They ran A/B tests proving that the AI output takes longer to fix than building from scratch. Leadership ignored the data. Oh, wait, there's more... They work in healthcare. You know... life-or-death training. And leadership is rushing AI avatars and coaches into production without proper vetting. This is a recipe for loss of life and lawsuits. The comments in the thread ranged from solidarity to frustration to fear. One person quoted Cory Doctorow: "AI can't do your job, but an AI salesman can convince your boss to fire you and replace you with an AI that can't do your job." (BTW... if you want some non-tech bro views on AI, read Cory's blog - Plauralistic) What they described is organizational malpractice, not AI adoption. Everything they mentioned: the ignored A/B AI slop test data, the mandates without expertise, the silenced dissent, the rushed deployment in high-stakes situations -- these aren't features of proper "AI transformation." They're symptoms of leadership failure dressed up in tech-forward language. I've spent the last three years deep in generative AI, researching, experimenting, teaching, and yes, I'm genuinely excited about what AI can do when used thoughtfully. Nothing makes me more upset than watching executives who've never built a single learning experience treat AI as a magic button for headcount reduction. Your expertise isn't obsolete. It's more necessary than ever. The irony of the current moment is that AI tools are only as good as your judgment guiding them. Every hallucination, every "overly wordy fancy infographic with zero engagement," every inaccurate piece of content that slips through -- these are failures of the process, not proof that the process no longer needs skilled designers. The companies mandating AI-first while ignoring their own IDs are going to learn this the hard way: morale drop, talent drain, novice-expert separation, and a dead-end culture. To everyone feeling demoralized: I understand. I talk to IDs every week who feel exactly what you're feeling. Years of education and craft are being swept aside. Some orgs are doing AI well and elevating their designers into more strategic roles. Others are creating disasters. One approach is sustainable. The other isn't. The ROI isn't there for slop. The lawsuits will come. Employee engagement will tank. Customers will notice. We need better leadership than what was described. And for what it's worth, this ID cared about quality; they ran A/B tests and worried about the healthcare implications -- this is exactly what makes IDs irreplaceable. No AI has that.

"If L&D only trains people to use AI tools, it misses its most strategic opportunity - shaping how people think and decide." Welcome again to the "Psychologist in the Boardroom" Series. #5 The biggest risk of AI adoption isn’t incompetence, it's the absence of judgment. Human judgment is a business-critical capability and L&D has a much bigger role to play than most organizations realise. Every ad on my Youtube, IG, even Prime Video these days is shortcuts to learning AI tools. Everyone is focused on training employees to use AI. Leaders have rolled out usage policies, adoption metrics are steadily rising - and yet, a quieter—far more serious—problem is emerging- people are using AI without thinking. AI outputs are accepted as neutral truth Decisions are deferred to algorithms without challenge. When decisions go wrong, I’m increasingly hearing, “That’s what the system suggested.” The problem is that tool training ≠ readiness to use. AI doesn’t eliminate human responsibility, it amplifies existing cognitive habits. Some of the most common ones being - Automation bias: Trusting machine outputs over human reasoning Cognitive offloading: Letting tools think for us, not just assist us False certainty: The more “intelligent” a system appears, the less people feel the need to question it. Modern L&D systems need to shift from completion rates and rollouts to developing critical thinking skills. Today’s employees must question AI outputs and balance efficiency with ethical and contextual judgment. You cannot build these capabilities through slide decks and quizzes. You start by defining it. I believe that when working with AI, judgment should comprise of 3 components - 1. Critical Evaluation The ability to ask: does this output make sense in this context? What might it be missing? 2. Decision-Making Under Ambiguity Comfort with uncertainty, trade-offs, and imperfect data—without freezing or defaulting to the tool. 3. Ethical & Contextual Reasoning Understanding consequences, bias, and accountability beyond policies and checklists. Organizations that are ahead of the curve are quietly shifting from- Tool demonstrations → Decision simulations Knowledge checks → Judgment reviews “Right answers” → Reasoned trade-offs Instead of asking: “Did you follow the AI recommendation?” They ask: “Why did you accept or override it?” Learning becomes less about correctness—and more about thinking quality. Leaders now need to- Model skepticism without cynicism Encourage questioning without slowing execution Own decisions - even when AI is involved If leaders default to “the system knows best,” employees will too. Judgment is culturally learned before it is formally trained. And this is why L&D is now a more fundamental need than ever before. Would you agree? #LearningAndDevelopment #FutureOfWork #HumanSkills #WorkforceTransformation #AIandPeople #OrganizationalDevelopment

AI skills for 2026! What will actually matter based on skills! Many people think learning AI means learning new tools. However tools change every few months. We have big questions! What truly creates an advantage? How we can think off and how we work with AI. Here’s what I believe will matter most in 2026: 1. AI Fluency & SMART PROMPTing! It’s not about typing random questions. It’s actually about knowing how to guide AI. Moreover, understanding what AI can (and cannot) do! Asking clear, structured prompts and turning a vague idea into a concrete outcome! Example: Instead of saying, *Write a report* say: “Summarize this data in 5 executive level bullet points, highlight risks, and suggest 3 action items. That’s AI fluency. 👏🏻👏🏻👏🏻👏🏻 2. Data-Driven Thought Process!! When AI gives outputs, smart professionals validate them, cleaning up messy data & identifying patterns, comparing options logically! As a result, Making decisions backed by insights!! 👏👏👏👏 Example: Use AI to analyze customer complaints, but we identify which trend impacts revenue most. AI supports. We decide.... 3. Automation == Leverage! Old story to revamp our business thru Technology: Real productivity starts when we stop doing repetitive tasks manually. Automating reports, Drafting standard emails, Generating meeting summaries Creating reusable workflows!!👏🏻👏🏻👏🏻👏🏻 Example: Imagine saving 1 hour daily through automation. That would be 250+ hours a year invested back into strategy. 4. Content, Research & Insight! AI speeds up creativity however quality still depends on us. Creating interactive presentations, Market research summaries and Visual ideation!👏👏👏👏 Example: AI drafts 80%. We refine tone, add experience, and inject authenticity and present in interactive way. That’s the winning formula. 👏🏻👏🏻👏🏻👏🏻👏🏻 5. Ethics & Integration! Using AI responsibly will become a core leadership skill to grow & flourish! As a word of caution is knowing where AI should NOT be used! Esp- Protecting data privacy, Integrating AI into workflows not replacing thinking and explaining AI’s value to stakeholders clearly thus companies don’t need AI hype. They need AI with accountability. 👏👏👏👏 6. Human with AI Collaboration! The future is not AI vs Humans. It’s AI + Humans. (Incremental) AI → speed, scale, structure!!!!! Humans → judgment, empathy, strategy!!!! The best professionals will know: When should AI lead? When should humans take control? Time to build leverage: combining intelligence (AI) with wisdom (human judgment). #AISkills #AIGeneration #AIwithHumans #IntelligenceWithwisdom!#ArtificialIntelligence #AI #GenerativeAI #AI2026 #FutureOfAI #AIFuture #MachineLearning #AIAutomation #AITransformation

The technology isn’t the end it’s the start. Whose jobs is it to tranform when the whole company just because citizen no code data scientists?

Stop saying 95% of AI projects fail. Start looking at HOW AI changed everyone's job. "AI projects fail" is a problem you can outsource. Pick a better model, find a better use case, run another pilot, fire/hire the SI. But having to look inwards and honestly say "we have no idea what AI did to our workforce" is a leadership challenge that compounds every quarter that question goes unanswered. BCG surveyed 1,000+ C-level executives and found 70% of AI implementation challenges are people- and process-related. Not technology. Not algorithms. McKinsey found that companies that intentionally redesign workflows around AI are nearly 3x more likely to achieve meaningful business impact. Deloitte's 2026 survey of 3,200+ respondents found 84% of organizations have not redesigned a single role for AI capabilities. The tools are deployed. The organization is untouched. This week, HBR published an 8-month ethnographic study of the human-level impacts of AI. Researchers embedded at a 200-person tech firm and found that AI didn't reduce workloads. It intensified them. As AI eroded capability boundaries, employees started picking up tasks that were never theirs. The lines of what was and wasn't their job began to blur. The buffer between work and rest started to erode, as no one knew if doing more faster meant going home earlier or filling the gap with more work. Multi-tasking became the normal operating mode as people ran multiple AI threads at once, stretching prioritization and focus to the breaking point. Because leadership hadn't taken it upon themselves to define what the new AI-led workplace means on a human level, the constraints the roles were designed around were gone. Without that definition, work doesn't shrink. It compounds. Look at it practically. There's Ted. Ted is a technical copywriter with 20 years experience. Ted now has AI. What is Ted's job? Is he writing, or editing what AI drafts? If his core skill just shifted from creation to judgment, are we measuring him on output or on quality of decisions? Does he still need the same reporting structure, or can he run autonomously with AI as his first-pass reviewer? Does his manager become a strategist instead of an editor? Can three Teds do what five did, and if so, what do the other two become? Every Fortune 500 has ten thousand Teds. Almost none of them are asking these questions. They're issuing AI subscriptions and counting adoption metrics while the actual work stays frozen in place. The roles, the scope, the reporting lines, the success criteria. All untouched. The priority AI question facing enterprise leaders shouldn't be whether their AI pilot worked. It's whether anyone can tell Ted what his job is on Monday morning.

Most professionals know they are under-utilizing AI. They know better prompting = better results, but they don't have time to figure it out. It's easier than you think. I’ve spent the last few weeks diving deeper and deeper down the AI rabbit hole. It started with redesigning our website with Google Gemini and AI Studio. Since then I’ve graduated to other tools, such as NotebookLM and Google Gems. I’ve started creating AI agents that are trained on the context of my business and actually increase productivity. The one constant throughout, and something most people struggle with: prompting. Whether you’re asking ChatGPT to summarize a PDF, or using Claude Code to build a full-stack web application, prompting is how we, the humans, interact with AI, the machines. And after consuming hours of AI content on YouTube and Twitter/X, I’ve learned that good prompting isn’t a new skill, but rather a new application of a skill most of us already have: leadership. You’re leading a machine instead of a human, but the trick is to treat them exactly the same. Here are three good leadership habits that turn generic prompting into precise results: 1️⃣ Provide Context Good leaders don’t just ask their employees to accomplish a task. They tell them why the task is important and what problems it will solve. Instead of starting your prompt with the task (“Build me a website”), start with the context (“I lead a small actuarial consulting practice specializing in Claims Cost Management and we need to improve our website”). Chatbots are generalists; give them context to make them specialists. 2️⃣ Create Clarity Good leaders know that communication is key to getting the right result. They ask questions like “Does what I’m asking make sense?”, or “Do you have questions on what we’re trying to achieve?”. Before asking AI to go to work, first ask it to interview you. It will ask you questions to gain clarity on your intended purpose, and your outputs will improve dramatically. 3️⃣ Specify Outputs Good leaders are specific in their instruction. They avoid vagueness by saying things like “we need an Excel tool showing CPT-level detail for Chemotherapy”, or “we need a PPT to present our new forecasting model to leadership.” Take the same approach with AI. Tell it exactly what you’re looking for. Better yet, show it examples of what you’re looking for. It doesn’t know what good looks like until you tell it. -- Getting good at prompting feels like a technical challenge, but it’s actually just leadership repackaged. You don’t need to master a new skill.  You just need to stop treating AI like a glorified Google search and start treating it like a high-performing member of your team.

Adoption of generative AI in communications is gaining ground, but confidence in the tools we are using is not. And, honestly, I get it. The very first time I used an AI tool, back in 2023, the experience wasn't exactly a slam dunk. I was facing multiple deadlines and needed to write a research-based article for a freelance client. Deciding to check out the new (and highly-touted) GPT capabilities, I put in what I now know to be a poorly written prompt for an article outline. What I got back was a full draft that looked so clean, so polished, so “done,” that I was stunned. In my new user giddiness and naivety, I opted to change just a few words to match the client’s style, attached it, and sent it to my agency counterpart feeling, well, pretty impressed and proud of myself. But -- it didn’t pass the quality review. Bad quotes, fake facts. Things we now know to watch for and check for. Thankfully, that draft never reached the client. I ended up having to redo the work I should have done from the beginning. Hour spent finishing the research, structuring the story, editing and fact-checking. And I learned a valuable lesson: these are tasks that AI can support, but they must be checked. AI cannot replace me, my judgment, or my understanding of my client and their market. Our human judgment is both the safety net and the secret sauce. And now, as I've connected with communicators across the country -- learning that the number of communicators using AI has increased to nearly 75% last year -- we are starting to experience these same moments of curiosity, surprise, frustration and ultimately understanding. Because skill‑building isn’t just about learning prompts or tools. It’s about learning how to think with AI — not blindly trust it. If your team is building AI literacy, you’re not alone. And if some days feel overwhelming, you’re definitely not alone. We’re all learning this together, one draft, one experiment, one quality‑check at a time.